English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In this research, an integrated detection model, Swin-transformer-YOLOv5 orSwin-T-YOLOv5, was proposed for real-time wine grape bunch detection to inheritthe advantages from both YOLOv5 and Swin-transformer. The research wasconducted on two different grape varieties of Chardonnay (always white berryskin) and Merlot (white or white-red mix berry skin when immature; red whenmatured) from July to September in 2019. To verify the superiority ofSwin-T-YOLOv5, its performance was compared against several commonlyused/competitive object detectors, including Faster R-CNN, YOLOv3, YOLOv4, andYOLOv5. All models were assessed under different test conditions, including twodifferent weather conditions (sunny and cloudy), two different berry maturitystages (immature and mature), and three different sunlightdirections/intensities (morning, noon, and afternoon) for a comprehensivecomparison. Additionally, the predicted number of grape bunches bySwin-T-YOLOv5 was further compared with ground truth values, including bothin-field manual counting and manual labeling during the annotation process.Results showed that the proposed Swin-T-YOLOv5 outperformed all other studiedmodels for grape bunch detection, with up to 97% of mean Average Precision(mAP) and 0.89 of F1-score when the weather was cloudy. This mAP wasapproximately 44%, 18%, 14%, and 4% greater than Faster R-CNN, YOLOv3, YOLOv4,and YOLOv5, respectively. Swin-T-YOLOv5 achieved its lowest mAP (90%) andF1-score (0.82) when detecting immature berries, where the mAP wasapproximately 40%, 5%, 3%, and 1% greater than the same. Furthermore,Swin-T-YOLOv5 performed better on Chardonnay variety with achieved up to 0.91of R2 and 2.36 root mean square error (RMSE) when comparing the predictionswith ground truth. However, it underperformed on Merlot variety with achievedonly up to 0.70 of R2 and 3.30 of RMSE.

Swin-transformer-yolov5 For Real-time Wine Grape Bunch Detection